USING MODERN FIRES TO ESTIMATE CHARCOAL AGE INHERITANCE AT PALEOSEISMIC SITES IN CALIFORNIA

The age of past earthquakes is typically calculated from radiocarbon-dated sequences of late Holocene deposits. Radiocarbon dates on charcoal provide a maximum age for the deposits because the age of plant material burned, the frequency of fire, and the rapidity of transport for detrital pieces make the charcoal age older than the deposit age. This age difference, known as the “inherited age” can be tens to hundreds of years but most studies assume it is short and do not formally address the impact on event dating. To quantify inherited age in the California Transverse Ranges, we dated charcoal deposits from two recent fires that burned at two paleoseismic sites along the San Andreas Fault (the Grand Fire in 2013 at Frazier Mountain and the Bobcat Fire in 2020 at Pallett Creek). The genus (e.g., Quercus, Arctostaphylos) or plant part (grass, leaf, bark) were identified for each of 53 dated samples prior to dating. We find that (1) inherited ages are the same at both sites despite differences in catchment size and fire histories, (2) short-lived or annual plant parts (grasses, leaves) or composite samples of fine charcoal are typically younger than twig or wood samples from longer-lived genus, and (3) combining all dates, the average charcoal age precedes the date of the deposit by ~25 cal years (300-year 95% range).

We next use the distribution of inherited ages as an empirical prior in Bayesian modeling of late Holocene deposits within OxCal (Bronk Ramsey, 2021). First, we correct the inherited age distribution to account for larger radiocarbon calibration uncertainties for dates prior to 1955 CE. The corrected inherited age distribution has mean age of ~90 cal years (470-year 95% range). Second, we examine how using this corrected distribution impacts layer and earthquake ages in a synthetic stratigraphic model. We find that layer and earthquake ages are younger when the prior is applied, but rather than a static offset, the effect is variable and limited by the amount of time between dated layers. Future efforts will examine the effect on existing paleoseismic sites such as Pallett Creek and Frazier Mountain.